Deregulation of the P14arf/MDM2/P53 Pathway Is A

[CANCER RESEARCH 60, 417–424, January 15, 2000] Deregulation of the p14ARF/MDM2/p53 Pathway Is a Prerequisite for Human 1 Astrocytic Gliomas with G1-S Transition Control Gene Abnormalities Koichi Ichimura, Maria Bondesson Bolin,2 Helena M. Goike, Esther E. Schmidt, Ahmad Moshref, and V. Peter Collins3 Department of Pathology, Division of Molecular Histopathology, University of Cambridge, Addenbrooke’s Hospital, Cambridge CB2 2QQ, United Kingdom [K. I., E. E. S., V. P. C], and Ludwig Institute for Cancer Research and Unit of Tumorpathology, Department of Oncology and Pathology, Karolinska Hospital, 171 76 Stockholm, Sweden [K. I., M. B. B., H. M. G., E. E. S., A. M., V. P. C.]

ABSTRACT There are several lines of evidence to indicate that deregulation of

G1-S transition control alone may be detrimental for cell survival: (a) Deregulation of G -S transition control in cell cycle is one of the 1 transfection of rodent cells with the adenoviral E1A and E1B genes important mechanisms in the development of human tumors including resulted in transformation due to the viral proteins binding to and astrocytic gliomas. We have previously reported that approximately two- inactivating pRB and p53. Transfection with E1A alone induced p53 thirds of glioblastomas (GBs) had abnormalities of G1-S transition control either by mutation/homozygous deletion of RB1 or CDKN2A (p16INK4A), dependent apoptosis (10); (b) when E2F1 was ectopically expressed in or amplification of CDK4 (K. Ichimura et al., Oncogene, 13: 1065–1072, rodent embryo fibroblasts, although the cell indeed entered S phase

1996). However, abnormalities of G1-S transition control genes may in- (11), apoptosis was induced in a p53-dependent manner (12); (c) loss duce p53-dependent apoptosis in cells. Recent investigations suggest that of pRB function leads to inappropriate progression through S phase ARF p14 is induced in response to abnormal cell cycle entry and results in and induces apoptosis in developing mouse lens fiber cells, which can p53 accumulation by inhibiting MDM2-mediated transactivational silenc- be overcome by simultaneous loss of p53 (13) or E2F1 (14). This ing and degradation of p53. To investigate the roles of the G -S transition 1 evidence suggests that, at least in some cell types, p53 prevents cells control system and the p14ARF/MDM2/p53 pathway in the development of astrocytic gliomas, we examined abnormalities of genes involved in these with deregulated G1-S control from abnormal proliferation by induc- regulatory pathways in a total of 190 primary human astrocytic gliomas of ing apoptosis. different malignancy grades [136 GBs, 39 anaplastic astrocytomas (AAs) p53 is a key regulator of cell cycle checkpoints. p53 binds to DNA and 15 astrocytomas (As)]. Sixty-seven percent of GBs (91/136) and 21% in a sequence-specific manner and functions as a transcription factor. of AAs (8/39) had abnormalities of the G1-S control system either by It induces either G1 arrest or apoptosis in response to various forms of mutation/homozygous deletion of RB1, CDKN2A or CDKN2B, or ampli- cell stresses (reviewed in Ref. 15). Expression of the p53 protein is fication of CDK4. Seventy-six percent of GBs (103 of 136), 72% of AAs (28 mainly regulated posttranscriptionally and maintained at a very low of 39), and 67% of As (10 of 15) had deregulated p53 pathway either by level in normal cells. MDM2 is an important regulator of p53. MDM2 mutation of TP53, amplification of MDM2, or homozygous deletion/mutation of p14ARF. When all of the data were combined and compared, 96% binds to p53 and inhibits its function by concealing the activation domain of p53 (16, 17) and by promoting degradation of p53, most of GBs (87 of 91) and 88% of AAs (7 of 8) with abnormal G1-S transition control also had deregulated p53 pathway. Thus, we demonstrate that likely through the ubiquitin-proteasome pathway (18, 19). In response deregulation of the G1-S transition control system was almost always to DNA damage, the MDM2 binding site of p53 is phosphorylated, accompanied by inactivation of the p53 pathway, clearly illustrating the the p53-MDM2 interaction is attenuated and p53 accumulates rapidly, cooperative roles of these two systems in the development/progression of relieved from MDM2-mediated suppression (reviewed in Ref. 20). primary human astrocytic gliomas. MDM2 is also one of the transcriptional targets of p53 (21). Recent investigations have identified another important regulator in INTRODUCTION this pathway, p14ARF (human homologue of mouse p19ARF). The ARF Frequent alterations of genes coding for proteins involved in G1-S p14 protein is encoded by the CDKN2A/INK4A locus but is transition control have been reported in GBs,4 the most malignant distinct from the p16INK4A protein (22). p14ARF is encoded by the form of brain tumor in adults (1–4). Almost mutually exclusive unique exon 1␤ (E1␤) and exon 2 and 3 of p16INK4A, using an involvement of RB1/CDK4/CDKN2A(p16INK4A) has been reported in alternative reading frame. E1␤ is located between exon 1␣ of more than 60% of these tumors (5). The proteins coded by these genes CDKN2A (E1␣) and exon 2 of CDKN2B on 9p21 (23, 24). It has been all directly or indirectly control the phosphorylation of pRB and the shown that the p14ARF protein binds to the p53/MDM2 complex and release of the E2F1 transcription factor (reviewed in Ref. 6, 7). GBs inhibits MDM2-mediated degradation of p53, which indicates that [malignancy grade IV according to the WHO classification (8)] may p14ARF is an upstream regulator of p53 via MDM2 (25–28). There is arise de novo but also by progression from astrocytic tumors of a also evidence suggesting that p53 down-regulates expression of lower malignancy grade, such as the AA (malignancy grade III) and p14ARF, which would establish an autoregulatory feedback loop be- A [malignancy grade II (9)]. tween p53, MDM2, and p14ARF (29). The most striking finding is that E2F1 transcriptionally up-regulates expression of p14ARF (29, 30). It Received 7/14/99; accepted 11/12/99. has been shown that in p19ARF null mouse embryonic fibroblasts, The costs of publication of this article were defrayed in part by the payment of page accumulation of p53 and induction of apoptosis after introduction of charges. This article must therefore be hereby marked advertisement in accordance with ARF 18 U.S.C. Section 1734 solely to indicate this fact. E1A was attenuated (31). Thus, p14 seems to be a critical com- 1 Supported by grants from the Swedish Cancer Society, Stockholm’s Cancer Society, ponent in the scrutiny of proliferation signals by the p53 system (32). King Gustaf V. Jubilee Fund, Axel and Margaret Ax:son Johnsons Funds, Lars Hierta Foundation, the Funds of the Karolinska Institute, and CAMPOD. These findings prompted us to determine the status of genes in- 2 ARF Present address: Department of Cell and Molecular Biology, Medical Nobel Institute, volved in G1-S transition control and the p14 /MDM2/p53 pathway Karolinska Institute, 171 77 Stockholm, Sweden. in a large series of astrocytic gliomas. We found that almost all 3 To whom requests for reprints should be addressed, at Department of Pathology, Division of Molecular Histopathology, University of Cambridge, Addenbrooke’s Hospi- astrocytic gliomas with altered G1-S transition control genes also had tal, Box 235, Cambridge CB2 2QQ, United Kingdom. Phone: 44-1223-336072; Fax: abnormalities of the p14ARF/MDM2/p53 pathway genes. Our findings 44-1223-216980; E-mail: [email protected]. 4 The abbreviations used are: GB, glioblastoma; A, astrocytoma; AA, anaplastic A; indicated that disruption of the p53 pathway is virtually obligatory for DGGE, denaturing gradient gel electrophoresis; STS, sequence-tagged site. these tumors with G1-S transition control gene abnormality. 417

Table 1 Primers for DGGE analysis of TP53 mutation GC GC Exon clampa Forward primer clampa Reverse primer exon 2 TCCCCACTTTTCCTCTTGCAG A TTTTCGCTTCCCACAGGTCTC exon 3 GACCTGTGGGAAGCGAAAATTC A AAAAGAGCAGTCAGAGGACCAGG exon 4a GGTCCTCTGACTGCTCTTTTCACC A GGTAGGTTTTCTGGGAAGGGACAG exon 4b A TCACTGAAGACCCAGGTCCAGATG CAGGCATTGAAGTCTCATGGAAG exon 4c TCACTGAAGACCCAGGTCCAGATG A CAGGCATTGAAGTCTCATGGAAG exon 5ab GCCGTGTTCCAGTTGCTTTATC A GTCGTCTCTCCAGCCCCAGC exon 5bb C GGCCAAGACCTGCCCTGTGC GTCGTCTCTCCAGCCCCAGC exon 5cb GGCCAAGACCTGCCCTGTGC A GTCGTCTCTCCAGCCCCAGC exon 6b C GGCCAAGACCTGCCCTGTGC GCCACTGACAACCACCCTTA exon 7b ACAGGTCTCCCCAAGGCGCA B CAGTGTGCAGGGTGGCAAGTG exon 8b TGATTTCCTTACTGCCTCTTG A CATAACTGCACCCTTGGTCT exon 9 A CTAAGCGAGGTAAGCAAGCAGG AAACGGCATTTTGAGTGTTAGACTG exon 10 TTACTTCTCCCCCTCCTCTGTTG A GGAATCCTATGGCTTTCCAACC exon 11 CACAGACCCTCTCACTCATGTGATG A TGCTTCTGACGCACACCTATTG a A, CCCCACGCCACCCGACGCCCCAGCCCGACCCCCCCGCGCCCGGCGCCCCC; B, CCCCGCTCCCCGCCCCCCTCCCCGCCCCGCCCCTCGCCGCCCCGGAC; C, CGCCCGCCGCGCCCCGCGCCCGTCCCGCCGCCCCCGCC. b Modified from Hamelin et al.(37).

MATERIALS AND METHODS where T ϭ tumor DNA and N ϭ control normal DNA. Mutation Analysis of CDKN2B and p14ARF. Mutation of exon 1 and exon Tumor Materials. Fresh surgical specimens from patients’ tumors were 2ofCDKN2B was examined by direct sequencing of PCR amplified products dissected into several macroscopically homogeneous pieces and stored at as described (3). Mutation of E1␤ of p14ARF was analyzed by direct sequenc- Ϫ135°C for up to 5 years before DNA/RNA extraction. A portion of each ing of the PCR products that covered the entire coding region (PC1340: frozen tumor piece was histologically examined for diagnosis and evaluation TGCAGTTAAGGGGGCAGGAG, forward; and PC1343: TTATCTCCTC- of the tumor cell content. Three tumors (AA17, AA59, and A26) had an CTCCTCCTAG-CCTG, reverse; putative start codon was according to Refs. estimated tumor cell content of 60%. All of the others had a minimum of 70% 25 and 30). The PCR products were directly sequenced using the same set of and generally more than 90% of tumor cells. Each patient’s blood was primers using the ABI PRISM BigDye Terminator Cycle Sequencing Kit (PE Ϫ collected before surgery and stored at 20°C before DNA extraction. The Biosystems, Foster City, CA) on an ABI PRISM 373aXL and a 377 DNA study was approved by the ethics committee of the Karolinska Hospital. Sequencer (PE Biosystems). Mutation in exon 2 of p14ARF was assessed by A total of 190 astrocytic gliomas were included in the present investigation. reevaluating the previously analyzed sequences of exon 2 of CDKN2A in the All of the tumors had been used in previous studies (1–3, 5, 33, 34). The same tumor series (3, 5). Any nucleotide changes in exon 2 of CDKN2A that histopathological diagnosis was carried out strictly according to the WHO can cause amino acid changes in the p14ARF protein were documented. classification (8). All of the cases collected before publication of the second DGGE Analysis of the TP53 Gene. Mutation of the TP53 gene (the gene edition of WHO classification (8) were reviewed in detail and reclassified. A encoding the p53 protein) was analyzed using DGGE. The entire coding region new tumor number with the appropriate prefix representing the diagnosis (GB, of TP53 as well as intron sequences adjacent to exons (minimum of six bases) AA, or A) was assigned to each reclassified case. All of the diagnoses are were covered by the analysis with the primer pairs listed in Table 1. A stretch based on the histopathology of the case, not the tumor piece, although the of a GC-rich sequence (GC-clamp) was attached to one of the primers for each pieces chosen are representative for each case. pair as indicated (Table 1). Primers for exons 5–8 are modified from Hamelin Allelic Assessment. Allelic assessment of RB1, CDK4, CDKN2A, et al. (37). The details of DGGE conditions and the strategy of DGGE analysis CDKN2B, and MDM2 was done by densitometry on Southern hybridization will be presented elsewhere.5 Paired DNA from tumors and patients’ blood using a PhosphorImager and ImageQuant software (Molecular Dynamics, were aligned in 96-well microtiter plates at a concentration of 10 ng/␮l. Thirty Sunnyvale, CA). DNA/RNA extraction, probes, Southern blotting, hybridiza- to 50 ng of DNA were used as PCR templates. A heteroduplex of tumor and tion and PhosphorImager analysis have been described previously (1–3, 5, 35). blood PCR products was prepared for each sample to increase the sensitivity ␤ ARF A probe for E1 of p14 was generated by PCR using primer pair PC978 of analysis. The heteroduplex was made by mixing an equal volume of PCR (CGAGTGAGGGTTTTCGTGGTTC, forward) and PC977 (CGTTGTAAC- products from tumor and blood, denaturing at 94°C for 15 min, incubating at CCGAATGG-GAAGC, reverse), which amplify a genomic segment contain- 65°C for 1 h and then at room temperature for 3 h. Thirty-four ␮lofthe Ј ␤ ␤ inga3 portion (138 bp) of E1 and part of intron 1 (145 bp). The primer heteroduplex was mixed with 10 ␮l of loading buffer (0.25% bromo-phenol- ␤ sequences were chosen based on the published genomic sequence around E1 blue, 0.25% xylene cyanol, and 30% glycerol). The D GENE system (Bio-Rad, (23). STS markers WI-3306 or WI-7427 were used as control probes for allelic Hercules, CA) was used to perform DGGE. An appropriate amount of urea and ARF assessment of p14 . WI-3306 is located on 2q close to D2S121, D2S112 and formamide was mixed with 6.5% acrylamide and cast according to the man- D2S44 and WI-7427 is located on 21q between D21S1257 and D21S270. ufacturer’s recommendation. Gels were run at 60°C at 160V for an optimal [Chromosome 2 workshop Consensus Map, GDB: 4225469; An STS-Based length of time for each primer pair. The gels were then stained with SYBR Map of the Human Genome, Data Files Release 12 (36)]. For each tumor, the green I (FMC, Rockland, ME) and digitally documented by an EagleEye II allelic status on 2q and 21q was determined by RFLP analysis with D2S44 or system (Stratagene, La Jolla, CA). When an abnormal shift was detected, the microsatellite analysis with D2S121, D2S112, D21S1257, and D21S270 (data corresponding exon was amplified from both tumor DNAand the patient’s not shown). A control locus from a chromosomal region that did not have blood DNA and directly sequenced to confirm the presence of a somatic ␤ allelic imbalance was chosen for each case. To assess allelic numbers at E1 mutation. DNA sequencing was performed using the ABI PRISM Ready of p14ARF, densitometry on TaqI digested Southern blotting was carried out. ARF Reaction DyeDeoxy Terminator Cycle Sequencing Kit (PE Applied Biosys- The signal intensity of the bands corresponding to p14 and the control locus tems) on an ABI PRISM 373 DNA Sequencer (PE Applied Biosystems) as in tumor and control DNA was measured using the ImageQuant, version 3.3, described previously (3). software (Molecular Dynamics, Sunnyvale, CA). The number of alleles at Statistical Analysis. A ␹2 test was used to statistically evaluate an associ- p14ARF was determined by the following formula: ation between various genetic abnormalities and to compare the incidences of p14ARF͑T͒ abnormalities among different malignancy grades. When the expected frequency was less than five in any category, a Fisher’s exact test was applied Control locus͑T͒ Allele number ϭ 2 ϫ instead. p14ARF͑N͒ Control locus͑N͒ 5 K. Ichimura, manuscript in preparation. 418

Table 2 Incidence of G1-S transition control and p53 pathway gene abnormalities in astrocytic gliomas

G1-S transition control genes p53 pathway genes

G1-S p53 and G1- RB1 CDK4 CDKN2A CDKN2B control TP53 MDM2 p14ARF p53 S control Tumors Total muta/del amp mut/del mut/del genes mut amp mut/del pathway genes GB 136 19 (14%) 18 (13%) 54 (40%) 50 (37%) 91 (67%) 49 (36%) 11 (8%) 55 (40%) 103 (76%) 87 (64%) AA 39 0 3 (8%) 5 (13%) 3 (8%) 8 (21%) 26 (67%) 0 5 (13%) 28 (72%) 7 (18%) A 15 0 0 0 0 0 10 (67%) 0 0 10 (67%) 0 Total 190 19 (9%) 21 (11%) 59 (31%) 53 (28%) 99 (52%) 85 (45%) 11 (6%) 60 (32%) 141 (74%) 94 (49%)

a mut, mutation; del, homozygous deletion; amp, amplification.

RESULTS and 9 nonsense mutations. Three point mutations were found in intronic consensus splice junction sequences. Thirteen tumors had two G1-S Transition Control Genes. We have previously reported the mutations (9 GBs, 3 AAs, and 1 A) and one GB (GB27) had three status of the following G1-S transition control genes, CDKN2A/ INK4A mutations. The details of the TP53 mutations as well as the allelic p16 , CDK4, and RB1, for all of the tumors in the series (1–3, 5). status of the TP53 locus and expression of the p53 protein as assessed In addition, homozygous deletion and mutation of CDKN2B/p15INK4B by immunohistochemistry will be presented elsewhere.5 In summary, were determined for all of the cases in the present study. 49 GBs (36%), 26 AAs (67%), and 10 As (67%) had TP53 mutations The cumulated data showed the following results (see Table 2). (Table 2). Three GBs had homozygous deletion of RB1 and an additional 15 had Amplification of MDM2 was determined by densitometry on loss of one allele and mutation of the other allele of RB1. In addition, Southern blotting for all of the 190 tumors. A small part of the data one GB (GB185) had an aberrant transcript lacking exons 14–17 of has been described previously (1, 38). Eleven GBs (8%) had ampli- RB1. This case had loss of one allele at RB1 without a detectable fication of MDM2 (Table 2; Fig. 1B). No amplification was found mutation in the other allele.6 In all, 19 GBs (14%) had total loss of among AAs or As. wild-type RB1. No homozygous deletion or mutation of RB1 was The allelic status of E1␤ of p14ARF was assessed in the following found in the AAs or As. Eighteen GBs (13%) and three AAs (8%) had manner. E1␤ of p14ARF is located approximately 20 kb centromeric to CDK4 amplification. Fifty GBs and four AAs had homozygous dele- E1␣ of CDKN2A/p16INK4A, and telomeric and in the close vicinity to tion of CDKN2A, and an additional four GBs and one AA had loss of exon 2 of CDKN2B (23). When homozygous deletion of both one allele and mutation of the other allele of CDKN2A. In all, 54 GBs CDKN2A and CDKN2B was identified E1␤ of p14ARF was considered (40%) and 5 AAs (13%) had total loss of wild type of CDKN2A. Fifty as being homozygously codeleted. This cohomozygous deletion was GBs (37%) and three AAs (8%) had homozygous deletion of demonstrated in an analysis of 13 glioma cell lines in which p14ARF CDKN2B. No mutation of CDKN2B was identified. E1␤, CDKN2A and CDKN2B were specifically examined (data not The patterns of these G -S transition control gene abnormalities in 1 shown). On the basis of this principle, 48 GBs and 3 AAs were individual tumors were then compared (Table 3). Eighteen of 19 GBs considered as having homozygous deletion of E1␤ of p14ARF (Table 3). with RB1 mutation/homozygous deletion had neither CDK4 amplifi- The allelic status of E1␤ of p14ARF was specifically examined in 15 cation nor CDKN2A/CDKN2B homozygous deletion. One GB (GB47) cases using a genomic fragment containing part of E1␤ and the had mutation of both RB1 and CDKN2A. Eighteen GBs and three AAs with CDK4 amplification retained at least one wild-type allele of RB1, adjacent intron as a probe for Southern hybridization. These include CDKN2A, and CDKN2B, whereas one GB (GB28) had both amplifi- the cases with homozygous deletion/mutation of only CDKN2A cation of CDK4 and homozygous deletion of CDKN2A (Table 3). (GB14, GB184, GB36, GB236, GB177, AA87, and AA86) or only Forty-eight GBs and three AAs had homozygous deletion or mutation CDKN2B (GB199 and GB147) and the cases with abnormalities of of both CDKN2A and CDKN2B but no RB1 mutation or CDK4 RB1 or CDK4 without TP53 mutation or MDM2 amplification [GB47, amplification. Six GBs and two AAs had homozygous deletion of GB100, GB25, GB157, GB11, and AA7 (Table 3)]. As a result, 5 only CDKN2A but not CDKN2B. Two GBs (GB199 and GB147) had tumors (GB14, GB184, GB199, GB147, and AA87) were determined ␤ ARF homozygous deletion of CDKN2B but not CDKN2A. When all of to have homozygous deletion of E1 of p14 (Fig. 1C). For the ␤ these abnormalities were combined, 91 GBs (66%), 8 AAs (21%), and remaining 10 cases, the coding region of E1 was sequenced to ␤ no As had either total loss of wild type of RB1, CDKN2A,or determine mutations. No nucleotide changes in E1 were identified. ARF INK4A CDKN2B, or amplification of CDK4 (Tables 2 and 3). For mutation in exon 2 of p14 , CDKN2A/p16 sequences were ARF p53 Pathway Genes. We then examined abnormalities of TP53, reevaluated in the context of the alternative reading frame for p14 INK4A MDM2, and p14ARF, which are involved in the regulation of the p53 [see above and (3, 5)]. Among five CDKN2A/p16 mutations pathway. All of the 190 tumors in the series were examined for identified, four were located in the common exon 2. These four INK4A mutation of the TP53 gene in exons 2–11 by DGGE. Primers for the CDKN2A/p16 mutations also altered the amino acid sequence of ARF DGGE analysis were chosen, and DGGE conditions were optimized p14 (Table 3). ARF so that any mutation in the entire coding region as well as intron Profiles of p14 /MDM2/p53 pathway gene abnormalities in in- sequences at exon/intron borders (minimum of six bases in introns) dividual tumors were then compared. Eleven tumors (9 GBs and 2 ARF would be detected by at least one combination of the primer pairs. AAs) had both TP53 mutation and homozygous deletion of p14 . Examples of DGGE results are shown in Fig. 1A. In all, 100 mutations Three tumors (two GBs and one AA) had mutation in TP53 and exon ARF were identified in 85 tumors (Table 2). There were 72 missense 2ofp14 . No tumor with MDM2 amplification had either TP53 ARF mutations, 3 inframe deletions, 10 frame-shift deletions, 3 insertions, mutation or p14 deletion/mutation. When abnormalities of all of the three genes were combined, 103 GBs (76%), 28 AAs (72%), and

6 10 As (67%) had either mutation of TP53, amplification of MDM2,or H. M. Goike. Clonal selection for genetic abnormalities for the G1/S transition control and p53 pathways in human glioblastoma xenografts, manuscript in preparation. homozygous deletion or mutation of p14ARF (Table 2). 419

Table 3 The profile of G1-S transition control and p53 pathway gene abnormalities in individual tumors Only cases with G1-S transition control gene abnormalities are listed. Tumors with abnormalities of only G1-S transition control genes but not p53 pathway genes are underlined. Nomenclature of mutation is according to recommendations of the HUGO Nomenclature committee whenever applicable.(62)

G1-S transition control genes p53 pathway genes Tumor No. RB1 CDK4 CDKN2A CDKN2B TP53 MDM2 p14ARF GB105 N258del na n n R110del n GB153 R445X n n n R248Q n GB10 Q444X n n n S127Y n GB12 H483ins n n n S241F n GB130 IVS8ϩ1GϾC n n n N132del n GB15 DEL n n n R273H n GB167 L337ins n n n R282W n GB195 IVS10Ϫ1GϾA n n n Y220C n GB235 IVS6ϩ1GϾA n n n S303del n GB39 Q575del n n n V197del n GB44 Y728ins n n n L35ins; R175H n GB129 K319X n n n A138P; R248Q n GB49 V494ins n n n IVS4ϩ5GϾAn GB185 splicing ex14–17 n n n V173L; R273C n GB140 DEL n n n n AMP GB47 IVS11Ϫ2AϾC n R80X n n n P94Lb,c GB100 DEL n n n n n nb,c GB25 IVS12ϩ1GϾAn n n n n nb,c GB157 W75X n n n n n nb,c GB28 n AMP DEL DEL A276P n DEL GB168 n AMP n n Y220C n GB248 n AMP n n L111Q; R282W n GB13 n AMP n n R248Q n GB189 n AMP n n C242S n GB82 n AMP n n H193Y n GB154 n AMP n n E171del; H193Y n GB142 n AMP n n n AMP GB237 n AMP n n n AMP GB245 n AMP n n n AMP GB246 n AMP n n n AMP GB35 n AMP n n n AMP GB37 n AMP n n n AMP GB7 n AMP n n n AMP GB81 n AMP n n n AMP GB88 n AMP n n n AMP GB90 n AMP n n n AMP GB11 n AMP n n n n nb,c GB17 n n DEL DEL P190L n DEL GB16 n n DEL DEL H179Q n DEL GB166 n n DEL DEL R213Q n DEL GB182 n n DEL DEL R342del n DEL GB33 n n DEL DEL G266R n DEL GB55 n n DEL DEL IVS3Ϫ2GϾA n DEL GB138 n n DEL DEL C135F n DEL GB22 n n DEL DEL R273C n DEL GB1 n n DEL DEL n n DEL GB101 n n DEL DEL n n DEL GB137 n n DEL DEL n n DEL GB150 n n DEL DEL n n DEL GB175 n n DEL DEL n n DEL GB176 n n DEL DEL n n DEL GB188 n n DEL DEL n n DEL GB197 n n DEL DEL n n DEL GB21 n n DEL DEL n n DEL GB213 n n DEL DEL n n DEL GB227 n n DEL DEL n n DEL GB251 n n DEL DEL n n DEL GB40 n n DEL DEL n n DEL GB45 n n DEL DEL n n DEL GB69 n n DEL DEL n n DEL GB84 n n DEL DEL n n DEL GB89 n n DEL DEL n n DEL GB94 n n DEL DEL n n DEL GB97 n n DEL DEL n n DEL GB99 n n DEL DEL n n DEL GB133 n n DEL DEL n n DEL GB135 n n DEL DEL n n DEL GB102 n n DEL DEL n n DEL GB126 n n DEL DEL n n DEL GB136 n n DEL DEL n n DEL GB159 n n DEL DEL n n DEL GB170 n n DEL DEL n n DEL GB18 n n DEL DEL n n DEL GB2 n n DEL DEL n n DEL GB226 n n DEL DEL n n DEL a n, no homozygous deletion, no amplification, or no mutation; X, stop codon; AMP, amplification; DEL, homozygous deletion; del, deletion; ins, insertion; ex, exon; HUGO, Human Genome Oganization. b Allelic status was directly assessed by Southern hybridization. c p14ARF exon 1␤ (E1␤) was examined for mutation. 420

Table 3 Continued.

G1-S transition control genes p53 pathway genes Tumor No. RB1 CDK4 CDKN2A CDKN2B TP53 MDM2 p14ARF GB244 n n DEL DEL n n DEL GB32 n n DEL DEL n n DEL GB38 n n DEL DEL n n DEL GB50 n n DEL DEL n n DEL GB52 n n DEL DEL n n DEL GB56 n n DEL DEL n n DEL GB87 n n DEL DEL n n DEL GB3 n n DEL DEL n n DEL GB63 n n DEL DEL n n DEL GB36 n n H83Y n Y234X n A97Vb,c GB236 n n M52del n G199del n D67delb,c GB177 n n Y44X n R273C; T304ins n nb,c GB14 n n DEL n V147del; R158G n DELb GB184 n n DEL n n n DELb GB199 n n n DEL n n DELb GB147 n n n DEL n n DELb AA17 n AMP n n R273C n AA92 n AMP n n H179R n AA7 n AMP n n n n nb,c AA18 n n DEL DEL C275F n DEL AA49 n n DEL DEL n n DEL AA93 n n DEL DEL n n DEL AA86 n n N71K n K320del n L86Vb,c AA87 n n DEL n R273C n DELb

ARF Correlation of G1-S Transition Control Gene and p53 Pathway GB157) had no TP53 mutation, MDM2 amplification, nor p14 Gene Abnormalities. On the basis of the above findings, profiles of homozygous deletion/mutation.

G1-S control gene abnormalities and p53 pathway gene abnormalities Among the 21 cases with CDK4 amplification (18 GBs, 3 AAs), 7 in each individual tumor were compared. Among the 19 GBs with GBs and 2 AAs had TP53 mutation (Table 3; see GB154 in Fig. 1, A RB1 homozygous deletion or mutation, 14 tumors had TP53 mutations and B). Among them, one GB with both CDK4 amplification and and 1 tumor (GB140) had MDM2 amplification (Table 3; Fig. 1B). homozygous deletion of CDKN2A/CDKN2B (GB28) had both TP53 One GB (GB47) with both RB1 and CDKN2A mutation had a muta- mutation and homozygous deletion of p14ARF (Table 3). An additional tion of exon 2 of p14ARF (Table 3). Three GBs (GB100, GB25, and 10 GBs had amplification of MDM2 (see GB90 in Fig. 1B). Only two

Fig. 1. A, DGGE analysis using primer pairs for exon 5b and exon 7 of TP53 (see Table 1). GB154, GB44 (ex5b), GB12, and GB36 (ex7) have abnormal shifts besides the normal band (N). B, Southern hybridization using probes for MDM2, CDK4, RB1, and D2S44. Matched patients’ blood DNA (B) and tumor DNA (T) were loaded next to each other. The same TaqI blots were serially reprobed. GB140 had amplification of only MDM2 but not CDK4. It also had homozygous deletion of a 3Ј portion of RB1 (м exon 18; Ref. 5). Using a cDNA probe for RB1, we found that the bands corresponding to exons 14–16 had an allelic number of 1.4 by densitometry whereas the bands corresponding to exons 21–22 had an allelic number of 0.5, which indicated that the region including these two exons were homozygously deleted. GB154 had amplification of only CDK4 but not MDM2. This case had mutation of TP53 (see A). GB90 had amplification of both MDM2 and CDK4. D2S44 was used to control the loading of DNA between blood and tumor from each patient. C, Southern hybridization using probes for E1␤ of p14ARF, exon 2 of CDKN2A, exon 1 of CDKN2B, and WI-3306. All of the four probes were simultaneously hybridized to the blot. GB14 had homozygous deletion of p14ARF, CDKN2A but not CDKN2B. GB36 had hemizygous deletion of p14ARF, CDKN2A, and CDKN2B. It had mutation of exon 2 of CDKN2A (3) and exon 7 of TP53 (Table 3 and A). GB147 had homozygous deletion of p14ARF and CDKN2B but not CDKN2A. A 5.5-kb aberrant band was observed when probed with the CDKN2B exon 1 probe alone but not with the exon 2 probe, which indicated that the breakpoint of deletion lay between exon 1 and exon 2 of CDKN2B (data not shown). WI-3306 (2q) was used as a control locus.

421

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2000 American Association for Cancer Research. CELL CYCLE AND p53 PATHWAY GENES IN ASTROCYTOMAS tumors (GB11 and AA7) had no TP53 mutation, MDM2 amplifica- abnormalities of G1-S transition control genes may give a growth tion, nor p14ARF homozygous deletion/mutation (Table 3). advantage to the cell; however, simultaneous inactivation of the p53 Among the 48 GBs and 3 AAs in which E1␤ of p14ARF was pathway is a prerequisite for cell survival when such abnormalities are considered as being homozygously codeleted with CDKN2A/p16INK4A present (10, 12, 13). It has been shown that introduction of wild-type and CDKN2B/p15INK4B, 9 GBs and 1 AA had TP53 mutation (Table p53 into the glioma cell lines U251 and A172, which had mutation of 3). Among two GBs and one AA with homozygous deletion of only TP53 and homozygous deletion of CDKN2A, led to apoptosis (42– CDKN2A but not CDKN2B, two (GB14 and AA87) had both homozy- 44). Our results substantiate the hypothesis proposed as a result of gous deletion of E1␤ of p14ARF and TP53 mutation (see GB14 in Fig. such experiments using cultured cells. 1C), and one (GB184) had homozygous deletion of E1␤ of p14ARF but The two systems can be targeted by a single genetic event. not TP53 mutation. Among the five tumors with CDKN2A/p16INK4A CDKN2A, CDKN2B, and p14ARF reside in a small region on 9p21, and mutation (GB47, GB36, GB236, GB177, and AA86), GB36, GB236, they are frequently homozygously codeleted in diverse types of hu- and AA86 had mutation of both TP53 and exon 2 of p14ARF (see man tumors including astrocytic gliomas (2, 45). CDK4 and MDM2 GB36 in Fig. 1, A and C, and Ref. 3), GB177 had only TP53 mutation, are located in the same chromosomal region, 12q13–15, albeit with a and GB47 had only exon 2 mutation of p14ARF (described above). 4-Mb gap between them (46). Amplification of this region may be Both of the cases with homozygous deletion of only CDKN2B but not initiated as a single event encompassing both MDM2 and CDK4. Later CDKN2A (GB199 and GB147) had homozygous deletion of E1␤ of the amplicon may be rearranged to include only the genes necessary p14ARF (see GB147 in Fig. 1C). to target, thus, explaining the finding that all of the loci between When all of the genetic abnormalities described above were com- MDM2 and CDK4 are not amplified in all of the tumors when studied bined and compared, 87 GBs and 7 AAs had abnormalities of both in detail (47). Rearrangements of amplicons in tumor cells have been

G1-S transition control genes and p53 pathway genes (Tables 2 and 3). well documented. One example is the rearrangement of the amplified Among the tumors with G1-S transition control gene abnormalities, EGFR gene in some GBs producing an aberrant, constitutively acti- this corresponds to 96% of GBs (87 of 91) and 88% of AAs (7 of 8). vated, receptor (48). Among tumors with p53 pathway gene abnormalities, 84% of GBs (87 However, in a large proportion of the cases, mutations of TP53 of 103) and 25% of AAs (7 of 28) had deregulated G1-S transition (located on 17p13) and abnormalities of the G1-S transition control control (Table 3). The association between abnormalities of G1-S genes examined here (RB1, on 13q14; CDK4, on 12q13–15; CDKN2A control genes and p53 pathway genes was highly significant among and CDKN2B, on 9p21) must have occurred as independent and GBs and among all of the tumors considered together (␹2 test, nonsynchronized events, because no single genetic event that would P Ͻ 0.0001). involve these genes simultaneously can be conceived. The association Four GBs and one AA with either RB1 mutation (GB100, GB25, of RB1 mutations/deletions and TP53 mutations alone is statistically and GB157) or CDK4 amplification (GB11 and AA7) had no TP53 significant (P ϭ 0.00081). Therefore, the association between the ARF mutation, MDM2 amplification, nor p14 homozygous deletion/ deregulated p53 pathway and G1-S transition control identified in this mutation. All of the tumors with homozygous deletion of CDKN2A study is unlikely to be caused merely by the genetic proximity of some and/or CDKN2B or CDKN2A mutation had either TP53 mutation or of these genes but is rather a consequence of selection for specific p14ARF homozygous deletion, or both. Sixteen GBs, 21 AAs, and 10 combinations of abnormalities.

A grade II had TP53 mutation but none of the G1-S control gene The coinactivation of the two systems was found exclusively in abnormalities. MDM2 amplification was always associated with either GBs and some AAs, which indicates that the combination of these RB1 mutation or CDK4 amplification. Twenty-nine GBs, 10 AAs, and abnormalities contributes to a more aggressive biological tumor phe- 5 As had neither homozygous deletion/mutation of RB1, CDKN2A, notype. Inactivation of the p53 pathway alone, exclusively by muta- CDKN2B,orTP53, nor amplification of CDK4 or MDM2. tion of TP53, was observed in all of the malignancy grades. In fact, the In two patients, tumors from the first and second surgeries were frequency of TP53 mutation is higher in As and AAs than in the GBs studied. In one of them, the initial tumor (AA100) had two mutations (␹2 test, P ϭ 0.0224 and 0.0007, respectively; see Table 2). It has been in exon 8 of TP53 (R273C and T304ins). When the tumor recurred shown that As with mutation of TP53 are more likely to recur and and progressed (GB177), it acquired a mutation in E1␣ of CDKN2A progress as compared with those without TP53 mutation (49). On the (Y44X) while retaining the same TP53 mutations as AA100. basis of these findings, an attractive model of astrocytic glioma progression considers mutation of TP53 an early step and the acqui- DISCUSSION sition of G1-S transition control gene abnormalities a later step. A good example is given by one AA (AA100) in our series with TP53 The majority of genes analyzed in this paper have been extensively mutation that progressed to GB (GB177) 3 years after the first studied, in general individually. Frequent abnormalities of these genes operation. It had then acquired a mutation in E1␣ of CDKN2A in have been documented in a variety of human tumors (reviewed in Ref. addition to the TP53 mutation documented in the AA. 39–41). However, no study has documented in detail the alterations It is clinically well documented that GB may arise as a de novo in a series of genes coding for components of interacting cellular tumor or may progress from astrocytic gliomas of lower malignancy. control mechanisms as is presented here. This type of study is nec- Judging from clinical data the majority of GBs are de novo tumors. essary if we are to understand the cellular mechanisms that are The arrangement of the genes in the genome provides the basis for aberrant in a particular tumor type. Individual genetic abnormalities single events that can inactivate the two cellular control systems should be viewed as ways of disrupting cellular control systems and simultaneously, providing a credible explanation for the prevalence of not as simply abrogating the function of a single gene. de novo GBs.

Our findings demonstrate that in primary human gliomas with Abnormalities of genes involved in G1-S transition and the p53 altered G1-S transition control, the p53 pathway is almost always pathway are frequently found in different human tumors. In our series, inactivated. Normally, p53 is involved in preventing cells from un- an individual tumor usually has an alteration of only one component controlled proliferation and tumor formation by inducing either G1 from each regulatory system, which supports the idea that abnormal- arrest or apoptosis when the cell enters the cell cycle in a nonphysi- ities of each component may have approximately equivalent effects. ological manner. Cell biological experiments have suggested that For example, RB1 mutation and CDK4 amplification are completely 422

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2000 American Association for Cancer Research. CELL CYCLE AND p53 PATHWAY GENES IN ASTROCYTOMAS mutually exclusive in our tumors. Not a single case had both a TP53 all of the As in our series lack clear evidence of a deregulated G1-S mutation and MDM2 amplification (see Table 3). However, in some control system. Many tumors in this category have loss of one allele tumors there were apparently redundant genetic abnormalities. One at RB1 or CDKN2A, without detectable mutation in the other alleles explanation could be that certain genetic events involving more than (3, 5). Methylation of the promoter region of CDKN2A has been one gene may also include a neighboring gene, i.e., homozygous suggested to be an alternative mechanism to inactivate the gene (58). deletion or amplification. In cases that have both TP53 mutation and However, our previous analysis of a limited number of tumors showed p14ARF homozygous deletion, p14ARF may simply have been code- only a very low incidence of methylation, and this did not correlate to leted with CDKN2A or CDKN2B without giving an additional selec- expression (3). A preliminary mutation analysis of the promoter tive advantage. This is supported by our finding that such redundant region of CDKN2A failed to identify somatic mutations.7 Mutation of abnormalities were found when CDKN2A, CDKN2B,orp14ARF were codon 24 of CDK4 has been reported in familial melanoma kindreds involved. Another possibility is that abnormalities of the different and sporadic melanomas, and the mutated protein has been demon- components may not have an equivalent effect. Even different muta- strated to function as kinase, yet it is unable to bind to p16INK4A (59, tions of the same gene may have varying impact on the function of its 60). Mutation analysis of CDK4 exon 2, which contains codon 24, protein product. Additionally, it cannot be excluded that multiple failed to identify any mutation in this tumor series.7 Other components abnormalities in the same complex pathway could provide an addi- in the same pathway, for example, CDK6 or members of Cyclin Ds, tional growth advantage. could possibly be involved (42, 61) and remain to be studied. p14ARF (p19ARF in mouse) has an unusual feature. Using an alter- The molecular mechanisms of growth control in cells is undoubt- native reading frame, it is translated partially from common exons edly more complex than our current understanding. The identification with CDKN2A/p16INK4A, yet results in a totally different protein of the molecular basis for the tumors without genetic abnormalities of (22–24). p14ARF has been found frequently deleted in a variety of the genes studied here will give further insight into tumorigenesis of human tumors, most often codeleted with CDKN2A but sometimes astrocytic brain tumors. specifically targeted (50). We demonstrated that in cases that have homozygous deletion of only CDKN2A or CDKN2B,E1␤ of p14ARF was always involved in the deletion, which indicated that E1␤ of ACKNOWLEDGMENTS p14ARF was the most frequently deleted region on 9p21. p14ARF We thank Lotta Asplund and Susanne O¨ hlin for excellent technical assist- knockout mice that specifically lack E1␤ developed normally but are ance and Alice Johnson-Marshall, Lu Liu, and Shohreh Varmeh-Ziaie for predisposed to various types of tumors including glioma (51, 52). ARF critical reading of the manuscript. We also thank the Departments of Neuro- These findings support p14 as a tumor suppressor gene. The surgery at Sahlgrenska Hospital, Go¨teborg, and the Karolinska Hospital, ARF ARF growth suppressive ability of p14 in p14 null human cells (or Stockholm, Sweden for their cooperation in the collection of clinical material. p19ARF in p19ARF null mouse cells) further supports this (28, 53, 54). Thus far no germ-line or somatic mutation in E1␤ has been identified in human tumors (see Ref. 52 and references therein). Mutations REFERENCES INK4A in exon 2 of CDKN2A/p16 often alter the amino acid sequence 1. Reifenberger, G., Reifenberger, J., Ichimura, K., Meltzer, P. S., and Collins, V. P. of p14ARF simultaneously, and four such mutations were identified in Amplification of multiple genes from chromosomal region 12q13–14 in human malignant gliomas: preliminary mapping of the amplicons shows preferential involve- this study. It has been reported that the NH2-terminal domain of ARF ␤ ment of CDK4, SAS, and MDM2. Cancer Res., 54: 4299–4303, 1994. p14 , exclusively coded by E1 , is necessary and sufficient for 2. Schmidt, E. E., Ichimura, K., Reifenberger, G., and Collins, V. P. CDKN2 (p16/ binding to MDM2 and inducing G1 arrest (27, 28, 54). The growth MTS1) gene deletion or CDK4 amplification occurs in the majority of glioblastomas. suppressive activity of p14ARF was not abrogated by a number of Cancer Res., 54: 6321–6324, 1994. 3. Schmidt, E. E., Ichimura, K., Messerle, K. R., Goike, H. M., and Collins, V. P. common mutations in exon 2 that alter the amino acid sequences of Infrequent methylation of CDKN2A(MTS1/p16) and rare mutation of both CDKN2A both p16INK4A and p14ARF (53, 54). These findings question the and CDKN2B(MTS2/p15) in primary astrocytic tumours. Br. J. Cancer, 75: 2–8, ARF 1997. significance of exon 2 mutations in p14 . Three of our four cases 4. Ueki, K., Ono, Y., Hensen, J. W., Efird, J. T., von Deimling, A., and Louis, D. N. with mutations in exon 2 of p14ARF also had a TP53 mutation. CDKN2/pl6 or RB alterations occur in the majority of glioblastomas and are inversely The role of CDKN2A/p16INK4A as a tumor suppressor gene has been correlated. Cancer Res., 56: 150–153, 1996. 5. Ichimura, K., Schmidt, E. E., Goike, H. M., and Collins, V. P. Human glioblastomas well established by both cell biological experiments and genetic with no alterations of the CDKN2A (p16INK4A, MTS1) and CDK4 genes have frequent analysis of primary human tumors in sporadic cases and in familial mutations of the retinoblastoma gene. Oncogene, 13: 1065–1072, 1996. tumor syndromes (55). Nevertheless, the mutation frequency of 6. Pines, J. The cell cycle kinases. Semin. Cancer Biol., 5: 305–313, 1994. 7. Weinberg, R. A. The retinoblastoma protein and cell cycle control. Cell, 81: 323–330, CDKN2A is relatively low compared with the frequency of homozy- 1995. gous deletion. It could simply reflect that simultaneous inactivation of 8. Kleihues, P., Burger, P. C., and Scheithauer, B. W. Histological Typing of Tumors of ARF the Central Nervous System, Ed. 2. New York: Springer-Verlag, 1993. both CDKN2A and p14 by homozygous deletion is a more effi- 9. Russell, D. S., and Rubinstein, L. J. Tumours of central neuroepithelial origin. In: cient way to simultaneously abrogate the p53 pathway and the G1-S Pathology of Tumours of the Nervous System, Ed. 5, pp. 83–350. London: Edward transition control. Arnold, 1989. 10. Debbas, M., and White, E. Wild-type p53 mediates apoptosis by E1A, which is It is intriguing that five cases had G1-S transition control gene inhibited by E1B. Genes Dev., 7: 546–554, 1993. abnormalities without TP53 mutation, MDM2 amplification or p14ARF 11. Johnson, D. G., Schwarz, J. K., Cress, W. D., and Nevins, J. R. Expression of homozygous deletion/mutation. The most likely explanation is that transcription factor E2F1 induces quiescent cells to enter S phase. Nature (Lond.), 365: 349–352, 1993. another component of the p53 pathway, either a regulating protein, for 12. Wu, X., and Levine, A. J. p53 and E2F-1 cooperate to mediate apoptosis. Proc. Natl. example, p300 (56), or downstream effector in the apoptosis pathway, Acad. Sci. USA, 91: 3602–3606, 1994. 13. Morgenbesser, S. D., Williams, B. O., Jacks, T., and DePinho, R. A. p53-dependent for example, BAX (57), may be altered in these cases. It is also apoptosis produced by Rb-deficiency in the developing mouse lens. Nature (Lond.), possible that some TP53 mutations have escaped detection, although 371: 72–74, 1994. we think that this is unlikely because primers and conditions for 14. Tsai, K. Y., Hu, Y., Macleod, K. F., Crowley, D., Yamasaki, L., and Jacks, T. Mutation of E2f-1 suppresses apoptosis and inappropriate S phase entry and extends DGGE were optimized to cover the entire coding region and splice survival of Rb-deficient mouse embryos. Mol. Cell, 2: 293–304, 1998. junction sequences. However, mutations in regulatory sequences lo- 15. Prives, C., and Hall, P. A. The p53 pathway. J. Pathol., 187: 112–126, 1999. cated outside of the screened region would naturally be missed. Approximately one-third of GBs as well as the majority of AAs and 7 K. Ichimura and V. P. Collins, unpublished data. 423

16. Momand, J., Zambetti, G. P., Olson, D. C., George, D., and Levine, A. J. The mdm-2 39. Greenblatt, M. S., Bennett, W. P., Hollstein, M., and Harris, C. C. Mutations in the oncogene product forms a complex with the p53 protein and inhibits p53-mediated p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. transactivation. Cell, 69: 1237–1245, 1992. Cancer Res., 54: 4855–4878, 1994. 17. Oliner, J. D., Pietenpol, J. A., Thiagalingam, S., Gyuris, J., Kinzler, K. W., and 40. Hall, M., and Peters, G. Genetic alterations of cyclins, cyclin-dependent kinases, and Vogelstein, B. Oncoprotein MDM2 conceals the activation domain of tumour sup- Cdk inhibitors in human cancer. Adv. Cancer Res., 68: 67–108, 1996. pressor p53. Nature (Lond.), 362: 857–860, 1993. 41. Momand, J., Jung, D., Wilczynski, S., and Niland, J. The MDM2 gene amplification 18. Haupt, Y., Maya, R., Kazaz, A., and Oren, M. Mdm2 promotes the rapid degradation database. Nucleic Acids Res., 26: 3453–3459, 1998. of p53. Nature (Lond.), 387: 296–299, 1997. 42. He, J., Allen, J. R., Collins, V. P., Allalunis-Turner, M. J., Godbout, R., Day, R. S., 19. Kubbutat, M. H., Jones, S. N., and Vousden, K. H. Regulation of p53 stability by III, and James, C. D. CDK4 amplification is an alternative mechanism to p16 gene Mdm2. Nature (Lond.), 387: 299–303, 1997. homozygous deletion in glioma cell lines. Cancer Res., 54: 5804–5807, 1994. 20. Prives, C. Signaling to p53: breaking the MDM2–p53 circuit. Cell, 95: 5–8, 1998. 43. Li, H., Lochmuller, H., Yong, V. W., Karpati, G., and Nalbantoglu, J. Adenovirus- 21. Piette, J., Neel, H., and Marechal, V. Mdm2: keeping p53 under control. Oncogene, mediated wild-type p53 gene transfer and overexpression induces apoptosis of human 15: 1001–1010, 1997. glioma cells independent of endogenous p53 status. J. Neuropathol. Exp. Neurol., 56: 22. Quelle, D. E., Zindy, F., Ashmun, R. A., and Sherr, C. J. Alternative reading frames 872–878, 1997. of the INK4a tumor suppressor gene encode two unrelated proteins capable of 44. Gomez-Manzano, C., Fueyo, J., Kyritsis, A. P., Steck, P. A., Roth, J. A., McDonnell, inducing cell cycle arrest. Cell, 83: 993–1000, 1995. T. J., Steck, K. D., Levin, V. A., and Yung, W. K. Adenovirus-mediated transfer of 23. Mao, L., Merlo, A., Bedi, G., Shapiro, G. I., Edwards, C. D., Rollins, B. J., and the p53 gene produces rapid and generalized death of human glioma cells via Sidransky, D. A novel p16INK4A transcript. Cancer Res., 55: 2995–2997, 1995. apoptosis. Cancer Res., 56: 694–699, 1996. 24. Stone, S., Jiang, P., Dayananth, P., Tavtigian, S. V., Katcher, H., Parry, D., Peters, G., 45. Hirama, T., and Koeffler, H. P. Role of the cyclin-dependent kinase inhibitors in the and Kamb, A. Complex structure and regulation of the P16 (MTS1) locus. Cancer development of cancer. Blood, 86: 841–854, 1995. Res., 55: 2988–2994, 1995. 46. Krauter, K., Montgomery, K., Yoon, S. J., LeBlanc-Straceski, J., Renault, B., Maron- 25. Stott, F. J., Bates, S., James, M. C., McConnell, B. B., Starborg, M., Brookes, S., del, I., Herdman, V., Cupelli, L., Banks, A., and Lieman, J. A second-generation YAC Palmero, I., Ryan, K., Hara, E., Vousden, K. H., and Peters, G. The alternative contig map of human chromosome 12. Nature (Lond.), 377 (Suppl): 321–333, 1995. product from the human CDKN2A locus, p14(ARF), participates in a regulatory 47. Reifenberger, G., Ichimura, K., Reifenberger, J., Elkahloun, A. G., Meltzer, P. S., and feedback loop with p53 and MDM2. EMBO J., 17: 5001–5014, 1998. Collins, V. P. Refined mapping of 12q13–q15 amplicons in human malignant gliomas 26. Pomerantz, J., Schreiber-Agus, N., Liegeois, N. J., Silverman, A., Alland, L., Chin, suggests CDK4/SAS and MDM2 as independent amplification targets. Cancer Res., L., Potes, J., Chen, K., Orlow, I., Lee, H. W., Cordon-Cardo, C., and DePinho, R. A. 56: 5141–5145, 1996. The Ink4a tumor suppressor gene product, p19Arf, interacts with MDM2 and neu- 48. Sugawa, N., Ekstrand, A. J., James, C. D., and Collins, V. P. Identical splicing of tralizes MDM2’s inhibition of p53. Cell, 92: 713–723, 1998. aberrant epidermal growth factor receptor transcripts from amplified rearranged genes 27. Zhang, Y., Xiong, Y., and Yarbrough, W. G. ARF promotes MDM2 degradation and in human glioblastomas. Proc. Natl. Acad. Sci. USA, 87: 8602–8606, 1990. stabilizes p53: ARF-INK4a locus deletion impairs both the Rb and p53 tumor 49. Reifenberger, J., Ring, G. U., Gies, U., Cobbers, L., Oberstrass, J., An, H. X., suppression pathways. Cell, 92: 725–734, 1998. Niederacher, D., Wechsler, W., and Reifenberger, G. Analysis of p53 mutation and 28. Kamijo, T., Weber, J. D., Zambetti, G., Zindy, F., Roussel, M. F., and Sherr, C. J. epidermal growth factor receptor amplification in recurrent gliomas with malignant Functional and physical interactions of the ARF tumor suppressor with p53 and progression. J. Neuropathol. Exp. Neurol., 55: 822–831, 1996. Mdm2. Proc. Natl. Acad. Sci. USA., 95: 8292–8297, 1998. 50. Kumar, R., Sauroja, I., Punnonen, K., Jansen, C., and Hemminki, K. Selective 29. Robertson, K. D., and Jones, P. A. The human ARF cell cycle regulatory gene deletion of exon 1␤ of the p19ARF gene in metastatic melanoma cell lines. Genes promoter is a CpG island which can be silenced by DNA methylation and down- Chromosomes Cancer, 23: 273–277, 1998. regulated by wild-type p53. Mol. Cell. Biol., 18: 6457–6473, 1998. 51. Kamijo, T., Bodner, S., van de Kamp, E., Randle, D. H., and Sherr, C. J. Tumor 30. Bates, S., Phillips, A. C., Clark, P. A., Stott, F., Peters, G., Ludwig, R. L., and spectrum in ARF-deficient mice. Cancer Res., 59: 2217–2222, 1999. Vousden, K. H. p14ARF links the tumour suppressors RB and p53. Nature (Lond.), 52. Kamijo, T., Zindy, F., Roussel, M. F., Quelle, D. E., Downing, J. R., Ashmun, R. A., 395: 124–125, 1998. Grosveld, G., and Sherr, C. J. Tumor suppression at the mouse INK4a locus mediated 31. de Stanchina, E., McCurrach, M. E., Zindy, F., Shieh, S. Y., Ferbeyre, G., Samuelson, by the alternative reading frame product p19ARF. Cell, 91: 649–659, 1997. A. V., Prives, C., Roussel, M. F., Sherr, C. J., and Lowe, S. W. E1A signaling to p53 53. Arap, W., Knudsen, E., Sewell, D. A., Sidransky, D., Wang, J. Y., Huang, H. J., and involves the p19ARF tumor suppressor. Genes Dev., 12: 2434–2442, 1998. Cavenee, W. K. Functional analysis of wild-type and malignant glioma derived 32. Pan, H., Yin, C., Dyson, N. J., Harlow, E., Yamasaki, L., and Van Dyke, T. Key roles CDKN2A␤ alleles: evidence for an RB-independent growth suppressive pathway. for E2F1 in signaling p53-dependent apoptosis and in cell division within developing Oncogene, 15: 2013–2020, 1997. tumors. Mol. Cell, 2: 283–292, 1998. 54. Quelle, D. E., Cheng, M., Ashmun, R. A., and Sherr, C. J. Cancer-associated 33. Ichimura, K., Schmidt, E. E., Miyakawa, A., Goike, H. M., and Collins, V. P. Distinct mutations at the INK4a locus cancel cell cycle arrest by p16INK4a but not by the patterns of deletion on 10p and 10q suggest involvement of multiple tumor suppressor alternative reading frame protein p19ARF. Proc. Natl. Acad. Sci. USA, 94: 669–673, genes in the development of astrocytic gliomas of different malignancy grades. Genes 1997. Chromosomes Cancer, 22: 9–15, 1998. 55. Serrano, M. The tumor suppressor protein p16INK4a. Exp. Cell Res., 237: 7–13, 1997. 34. Liu, L., Ichimura, K., Pettersson, E. H., and Collins, V. P. Chromosome 7 rearrange- 56. Grossman, S. R., Perez, M., Kung, A. L., Joseph, M., Mansur, C., Xiao, Z. X., Kumar, ments in glioblastomas; loci adjacent to EGFR are independently amplified. J. Neu- S., Howley, P. M., and Livingston, D. M. p300/MDM2 complexes participate in ropathol. Exp. Neurol., 57: 1138–1145, 1998. MDM2-mediated p53 degradation. Mol. Cell, 2: 405–415, 1998. 35. Ichimura, K., Schmidt, E. E., Yamaguchi, N., James, C. D., and Collins, V. P. A 57. Rampino, N., Yamamoto, H., Ionov, Y., Li, Y., Sawai, H., Reed, J. C., and Perucho, common region of homozygous deletion in malignant human gliomas lies between the M. Somatic frameshift mutations in the BAX gene in colon cancers of the microsat- IFN-␣/␻ gene cluster and the D9S171 locus. Cancer Res., 54: 3127–3130, 1994. ellite mutator phenotype. Science (Washington DC), 275: 967–969, 1997. 36. Hudson, T. J., Stein, L. D., Gerety, S. S., Ma, J., Castle, A. B., Silva, J., Slonim, D. K., 58. Merlo, A., Herman, J. G., Mao, L., Lee, D. J., Gabrielson, E., Burger, P. C., Baylin, Baptista, R., Kruglyak, L., Xu, S. H., Hu, X., Colbert, A., Rosenberg, C., Reeve-Daly, S. B., and Sidransky, D. 5Ј CpG island methylation is associated with transcriptional M. P., Rozen, S., Hui, L., Wu, X., Vestergaard, C., Wilson, K., Bae, J., Maitra, S., silencing of the tumour suppressor p16/CDKN2/MTS1 in human cancers. Nat. Med., Ganiatsas, S., Evans, C., DeAngelis, M., Ingalls, K., Nahf, R., Horton, L., Oskin, M., 1: 686–692, 1995. Collymore, A., Ye, W., Kouyoumjian, V., Zernsteva, I., Tarn, J., Devine, R., Court- 59. Wo¨lfel, T., Hauer, M., Schneider, J., Serrano, M., Wo¨lfel, C., Klehmann, H. E., De, ney, D., Renaud, M., Nguyen, H., O’Connor, T., Fizames, C., Faure, S., Gyapay, G., P. E., Hankeln, T., Meyer zum Bu¨schenfelde, K.-H., and Beach, D. A p16INK4a- Dib, C., Morissette, J., Orlin, J., Birren, B., Goodman, N., Weissenbach, J., Hawkins, insensitive CDK4 mutant targeted by cytolytic T lymphocytes in a human melanoma. T., Foote, S., Page, D., and Lander, E. An STS-based map of the human genome, with Science (Washington DC), 269: 1281–1284, 1995. supplementary data from the Whitehead Institute/MIT Center for Genome Research, 60. Zuo, L., Weger, J., Yang, Q. B., Goldstein, A. M., Tucker, M. A., Walker, G. J., Human Genetic Mapping Project, Data Release 11 (October 1996). Science (Wash- Hayward, N., and Dracopoli, N. C. Germline mutations in the p16INK4a binding ington DC), 270: 1945–1954, 1995. domain of CDK4 in familial melanoma. Nat. Genet., 12: 97–99, 1996. 37. Hamelin, R., Jego, N., Laurent, P. P., Vidaud, M., and Thomas, G. Efficient screening 61. Costello, J. F., Plass, C., Arap, W., Chapman, V. M., Held, W. A., Berger, M. S., Su of p53 mutations by denaturing gradient electrophoresis in colorectal tumors. Onco- Huang, H. J., and Cavenee, W. K. Cyclin-dependent kinase 6 (CDK6) amplification gene, 8: 2213–2220, 1993. in human gliomas identified using two-dimensional separation of genomic DNA. 38. Reifenberger, G., Liu, L., Ichimura, K., Schmidt, E. E., and Collins, V. P. Amplifi- Cancer Res., 57: 1250–1254, 1997. cation and overexpression of the MDM2 gene in a subset of human malignant gliomas 62. Antonarakis, S. E. Recommendations for a nomenclature system for human gene without p53 mutations. Cancer Res., 53: 2736–2739, 1993. mutations. Nomenclature Working Group. Hum. Mutat., 11: 1–3, 1998.

424

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2000 American Association for Cancer Research. Deregulation of the p14ARF/MDM2/p53 Pathway Is a Prerequisite for Human Astrocytic Gliomas with G 1-S Transition Control Gene Abnormalities

Koichi Ichimura, Maria Bondesson Bolin, Helena M. Goike, et al.

Cancer Res 2000;60:417-424.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/60/2/417

Cited articles This article cites 60 articles, 25 of which you can access for free at: http://cancerres.aacrjournals.org/content/60/2/417.full#ref-list-1

Citing articles This article has been cited by 26 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/60/2/417.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/60/2/417. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.